Invited or Not Invited? Does It Matter for Breast Cancer Mortality and Medical Expenditures?

Breast cancer is the leading cause of cancer-attributable mortality of females around the globe being responsible for 14% of such deaths. Screening mammography—a mammogram taken when there is no sign of breast cancer—is the recommended early detection tool for breast cancer due to its ability to detect pre-cancerous cells while they are still treatable. Following this recommendation many countries in the EU have initiated organized breast cancer screening programs where women in a targeted age range are invited for screening. The Netherlands operates such a publicly funded screening program covering all women within the target age range – 50-69 since 1989, extended to 50-75 in 1998. Women within the age interval get personal invitations every two years for a mammogram at no cost. In 2010, annual total cost of the program was 55 million Euros.

Despite the great amount of faith and money put into screening, nowadays the benefits are open to debate. While Gotzsche and Olsen (2005)[1] argue that there is no reliable evidence that screening reduces breast cancer mortality, Kalager et. al. (2010)[2] find that the reduction in breast cancer mortality due to screening is lower after taking advances in breast cancer treatment into account. With this study, we aim to contribute to this debate by evaluating the causal impact on breast cancer mortality and on health care expenditures of receiving an invitation from the Dutch breast cancer screening program.

We have access to Dutch administrative data for the 1995-2011 period, which contains accurate, individual level information on, among others, date and cause of death, medical expenditures, and residential address at the post code level. Our identification strategy relies on the geographical expansion of the program between 1989 and 1997 at the municipality level. This means that some municipalities implemented the program earlier than others, creating variation across women of same birth year in timing of their first invitations. Since expansion of the program occurred non-systematically, simply comparing breast cancer mortality rates and medical expenditures among women who are at the same age but differ in timing of the first invitation will give us the causal impact of the program on breast cancer mortality and medical expenditures.

Our findings show that the probability of dying from breast cancer is not lower among women who got invited by the screening program earlier. Similarly cumulative breast cancer mortality is not lower in municipalities which started the program earlier. Our very preliminary results on health care expenditures show that there is a positive effect of receiving an invitation on health care expenditures.

---

[1] Gotzsche, P.C. and Olsen, O. (2000), “Is Screening for Breast Cancer with Mammography Justifiable?”, Lancet 355, 129-34.

[2] Kalager, M., Zelen, M., Langmark, F. and Adami, H. (2010), “Effect of Screening Mammography on Breast Cancer Mortality in Norway”, The New England Journal of Medicine 363, 1203-10